Optimal Satellite Network Topology Design with Time-Dependent Traffic Demands

The use of satellite networks, either with radio frequency or optical links, emerges as a solution to tackle the digital divide problem. Oftentimes, the literature approaches the design of satellite networks using metaheuristic algorithms, which do not guarantee convergence to the global optimum. Furthermore, coverage is frequently adopted as the main fig-ure of merit, neglecting time-dependent traffic demands and the channel capacity of the links during network design. In this paper, we propose a mixed-integer linear programming formulation that simultaneously determines the optimal satellite constellation configuration and its network topology to satisfy strict time-dependent traffic demands, leveraging the concept of traffic matrices. We present two case studies to demonstrate the applicability of the model by defining a heterogeneous set of nodes (i.e., gateways, fishing vessels) with time-varying traffic demands, radio frequency links between satellites and nodes, and optical inter-satellite links. Additionally, the case studies provide insights into the coupled relationship between the satel-lite constellation configuration and the network topology.